Zero-downtime migration, owned by our engineers
Eight moves, each built on the method the engine documents, run by OSStend engineers into our cloud, your cloud account or your data center.
45 minutes with an engineer: we review your workload and sketch the design.
- ✓Operated by AceMQ
- ✓Engine-documented methods
OSStend migrations use the method each engine documents: federation and shovel for RabbitMQ, logical replication for PostgreSQL, replication for Valkey, mirroring for Kafka, and reindexing or snapshots for OpenSearch. Engineers run the migration and own the cutover, with both sides live until you switch.
- PostgreSQL
- MySQL
- MongoDB
- Redis
- Valkey
- OpenSearch
- Kafka
- RabbitMQ
- AWS
- Google Cloud
- Azure
- Kubernetes
- OpenShift
- Prometheus
- Grafana
- Datadog
- Dynatrace
- Splunk
Pick the move you are planning
Each guide covers one move: the documented method, what to check before you start, the phases and the cutover. Where a method has a pause or a version limit, the guide says so and links to the engine's own documentation.
Amazon MQ to RabbitMQ
Federation, shovel and the RabbitMQ blue-green method, with the Amazon MQ plugin rules that shape the link.
Read the guidePostgreSQLAmazon RDS to PostgreSQL
Logical replication off RDS, what it does not copy, and where the cutover pause falls.
Read the guideValkeyRedis to Valkey
Replication from Redis OSS 7.2 and earlier, an RDB copy where it fits, and the versions that do not carry over.
Read the guideMongoDBMongoDB Atlas to self-managed MongoDB
mongosync for a live move, mongodump for an offline one, and the version limits of each.
Read the guidePostgreSQLOracle to PostgreSQL
ora2pg for schema and data, a change-capture tool for the live gap, and the PL/SQL review.
Read the guidePostgreSQLMySQL to PostgreSQL
pgloader for the load, Debezium for live changes, and what pgloader leaves behind.
Read the guideKafkaConfluent to Apache Kafka
MirrorMaker 2 for topics and consumer offsets, plus schemas and connectors.
Read the guideOpenSearchElasticsearch to OpenSearch
Remote reindex, snapshot restore or Migration Assistant, and the version limits that choose between them.
Read the guideThe engines themselves are on the managed PostgreSQL, Kafka and RabbitMQ hosting pages. The process is the same for every move.
Moving off a managed service
Most of these moves start from a managed service with documented limits. The matching alternatives pages set out those limits, with sources and dates, and say where the service is still the better choice.
Amazon MQ
Cluster size, plugin list and upgrade timing, quoted from AWS documentation.
Compare with Amazon MQAlternativesAmazon RDS
PostgreSQL and MySQL outside one cloud, and when RDS is the better fit.
Compare with RDSAlternativesAmazon ElastiCache
Redis or Valkey with the admin commands ElastiCache restricts.
Compare with ElastiCacheAlternativesRedis Enterprise
Redis or Valkey run around the workload, in any cloud or on-prem.
Compare with Redis EnterpriseAlternativesMongoDB Atlas
MongoDB in OSStend's cloud, your cloud account or your data center, run 24/7 by AceMQ engineers.
Compare with AtlasAlternativesConfluent
Apache Kafka with Connect and a schema registry, in your account or data center.
Compare with ConfluentAlternativesElastic Cloud
OpenSearch run around the workload, with the move planned and run for you.
Compare with Elastic CloudThe full list is on the alternatives page, including Heroku Postgres for teams leaving a platform database. Longer reads on choosing a provider are in the guides.
The eight moves side by side
The method decides how close a move gets to zero downtime. Replication-based methods keep the old and new side live until the switch, while a dump and restore needs a pause, so the design review picks a replication method wherever the engine documents one.
| Move | Documented method | Both sides live before the switch | Official docs |
|---|---|---|---|
| Amazon MQ to RabbitMQ | Definitions import, federation, shovel; the RabbitMQ blue-green method | Yes: federation and shovel carry messages | RabbitMQ docs |
| Amazon RDS to PostgreSQL | Logical replication; pg_dump and pg_restore for the schema | Yes with logical replication, no with a dump alone | PostgreSQL docs |
| Redis to Valkey | Replication with REPLICAOF; RDB snapshot copy | Yes with replication, no with a snapshot copy | Valkey docs |
| Atlas to self-managed MongoDB | mongosync; mongodump and mongorestore | Yes with mongosync until commit, no with mongodump | MongoDB docs |
| Oracle to PostgreSQL | ora2pg for schema and data; change capture for live changes | Only when a change-capture tool is added | ora2pg docs |
| MySQL to PostgreSQL | pgloader for the load; Debezium for live changes | Only when a change-capture tool is added | pgloader docs |
| Confluent to Apache Kafka | MirrorMaker 2 | Yes: it replicates continuously | Apache Kafka docs |
| Elasticsearch to OpenSearch | Remote reindex; snapshot restore; Migration Assistant | Yes with remote reindex or Migration Assistant; snapshot restore needs downtime or change capture | OpenSearch docs |
Sources: the engine documentation linked in each row and in each guide; checked 30 Oct 2026.
What owning the cutover means
We run the migration and own the cutover, with both sides live until you switch. Zero downtime is the plan.
Review
An engineer reviews the current setup: sizes, settings, versions, traffic and the limits the source service enforces.
Design
The new cluster and the migration method are written down and signed off before anything is built.
Link both sides
The documented method keeps the old and new side in step while both stay live.
Cut over
Traffic switches only when the new side is verified, and the old side stays until you are sure.
The timeline and runbook for each engine's migration are written in the design step, from your data volume, traffic and change windows, and you approve them before the cutover is scheduled.
Who runs the migration
AceMQ engineers run every move. AceMQ has operated open source data and messaging systems for 10+ years and serves 160+ customers, with engineers on call 24/7.
IBM MQ to RabbitMQ, without replacing the core
American National Insurance moved its Windows-based IBM MQ queues to RabbitMQ on Kubernetes first, with both brokers live until behaviour was validated, while IBM MQ licensing costs were progressively removed. Use case on acemq.com.
AceMQ engagementA Kafka CDC pipeline it did not need
American National replaced a Kafka-based change data capture pipeline for SQL Server with standalone Debezium publishing to RabbitMQ, removing Kafka clusters and Kafka Connect the use case did not need. Use case on acemq.com.
AceMQ engagementAceMQ names a customer only with that customer's written permission, so most of this record stays anonymous. Ask for a similar workload in the 45-minute review.
Questions about migration
Does every migration have zero downtime?
Zero downtime is the plan for each move, and the method decides how close it gets. Replication-based methods keep both sides live. A dump and restore needs a pause, so the design review picks a replication method wherever the engine documents one.
Can the migration be run without OSStend?
Yes. Each guide links to the engine's own documentation, and the methods work without OSStend. The guides also say where a move needs care, so a team that runs it alone can plan for it.
How long does a migration take?
Duration depends on data volume, change rate, the number of applications and how long both sides must stay live. An engineer gives an estimate for the specific workload during the design review.
Related pages
- Migration guideAmazon MQ to RabbitMQ migration without downtimeAmazon MQ to RabbitMQ migration without downtime: the federation, shovel and blue-green methods RabbitMQ documents, Amazon MQ's plugin rules and cutover.
- Migration guideMigrate Postgres off Amazon RDS, step by stepMigrate Postgres off Amazon RDS with logical replication and pg_dump: what to switch on, what replication skips, how cutover works and where the pause falls.
- Migration guideRedis to Valkey migration: what changes and howRedis to Valkey migration with replication or an RDB copy: which Redis versions carry over, what changes for clients and what to test before the switch.
- Migration guideMigrate MongoDB off Atlas without downtimeMigrate MongoDB off Atlas with mongosync or mongodump: version limits, what to check, how cutover works, and managed running by AceMQ engineers.
- Migration guideOracle to PostgreSQL migration with ora2pgOracle to PostgreSQL migration with ora2pg: what it converts, the PL/SQL work left by hand, how change capture keeps Oracle live, and how cutover works.
- Migration guideMySQL to Postgres: pgloader, psql, DBeaver and DMSHow to connect MySQL to PostgreSQL and move the data: pgloader commands, a tab file with psql \copy, DBeaver and AWS DMS, plus change capture and cutover.
Bring the move you are planning
45 minutes with an engineer: the current setup is reviewed against the documented method and the cutover is sketched. No ticket queue.
- ✓A design sketch for your workload, not a plan menu
- ✓A zero-downtime cutover path for what you run today
- ✓A price for that design, after the review